04194ntm a22005297i 4500 000723017 CZ-PrVSE 20250614131007.0 m d cr n|||||||||| 250614s2025 xr fsbm 000 0 eng d NEZPRACOVANÝ IMPORT ABA006 cze ABA006 ABA006 rda Fuka, Dominik ISIS:155645 dis Aplikace systému LISp-Miner a data mining technik na reálná podniková data eng Application of LISp-Miner system and data mining techniques on real-life business data / Dominik Fuka 2025 ?? stran : digital, PDF soubor Vedoucí práce: Jan Rauch Diplomová práce (Ing.)—Vysoká škola ekonomická v Praze. Fakulta informatiky a statistiky, 2025 Obsahuje bibliografii Textový (vysokoškolská kvalifikační práce) Rok obhajoby 2025 This thesis explores a practical application of the LISp-Miner system in combination with the CRISP-DM methodology to evaluate e-commerce data from Butterfly Playing Cards, s.r.o., a small Czech company. It operates two online stores - Butterfly Magic Store (BMS) and Butterfly Wonderland (BW) - each serving a different market (international and Czech). The aim of the thesis was to examine how could LISp-Miner help introduce data mining to a real-world business with limited resources for data analyst. The research followed the CRISP-DM structure. It went through phases of business understanding, data preparation, modeling, evaluation, and planning the deployment. Several analytical questions were formed in collaboration with the store manager to create the goals of the analysis. These questions focused on optimizing free shipping thresholds for EU and rest of the world, identifying frequently purchased product combinations, trying to uncover product preferences in different countries, and confirming seasonal sales patterns. Real-life datasets were exported from the Shoptet platform. Some preprocessing was done in Excel and Python before import to LISp-Miner. Then, the data was imported into LISp- Miner where attributes and tasks were created. CF-Miner, 4ft-Miner, SD4ft-Miner, and SDCF-Miner modules were utilized to generate findings. The analysis confirmed, among other insights, that a significant number of customers do not get motivated by free shipping thresholds, and that there are some opportunities for product bundles and marketing adjustments. Despite limitations of the thesis, such as incomplete or noisy data attributes, GDPR-related exclusion of attributes, and Shoptet export constraints, the thesis demonstrated that a transparent and replicable data mining process is possible to establish even in a small developing company. The results provide not only immediate business insights but also a reusable analytical framework. This work can contribute to the current discussion on the value of interpretable data mining in business decision-making. It can also offer a practical guide for similar organizations looking into leveraging their data. Způsob přístupu: Internet znalostní a webové technologie [obor dipl. práce] diplomové práce fd132022 czenas master's theses eczenas LISp-Miner GUHA data mining CRISP-DM business data Rauch, Jan, 1948- jn20001227021 ths Máša, Petr ISIS:17194 opn Vysoká škola ekonomická v Praze. Fakulta informatiky a statistiky kn20010709399 dgg https://insis.vse.cz/zp/89849/podrobnosti VŠKP v InSIS https://insis.vse.cz/zp/89849 Hlavní práce https://insis.vse.cz/zp/89849/posudek/vedouci Hodnocení vedoucího https://insis.vse.cz/zp/89849/posudek/oponent/86474 Oponentura https://insis.vse.cz/zp/89849/podrobnosti dc:identifier NEPOSILAT VSKP vse89849 250613 89849